Public Use Databases: A Precursor To Analysis Databases

• Activity Number: 34
• Type: Topic Contributed
• Date/Time: Sunday, August 11, 2002 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section*
• Author(s): James Hosking*+ and Russell Helms
• Affiliation(s): University of North Carolina, Chapel Hill and Rho, Inc.
• Keywords: Analysis Database ; Clinical Trials ; Public Use Database

Abstract

Since about 1980, after completion of publication activities for a large NIH-sponsored clinical trial, the coordinating center typically prepares and makes public a collection of analysis datasets called "public use datasets." Public use datasets are important in that they can be readily accessed and analyzed by diverse researchers. While public use datasets are useful, they are, in fact, collections of datasets rather than true databases, and thus may not facilitate integration of data from multiple sources and may require elaborate data manipulation to produce some types of analysis-ready files. In contrast, an Analysis Database is an advanced solution to integrating clinical trial data. This presentation describes the historical development and use of these datasets in the health sciences, existing NIH guidelines for creation of these collections of datasets, and their importance in the conceptualization and design of a compound analysis database for drug development. The drug development analysis database model is extended to address the issue of how public use datasets from individual trials can be synthesized into broader analysis databases for public health decision-making.